WO2015009020A1

WO2015009020A1 - Method and apparatus for encoding/decoding scalable video signal

Info

Publication number: WO2015009020A1
Application number: PCT/KR2014/006374
Authority: WO
Inventors: 이배근; 김주영
Original assignee: 주식회사 케이티
Priority date: 2013-07-15
Filing date: 2014-07-15
Publication date: 2015-01-22
Also published as: KR20150133684A; CN105379275A; KR20150009466A; US20160156913A1

Abstract

A method for decoding a scalable video signal according to the present invention comprises: obtaining a discardable flag for a picture in a lower layer; determining whether the picture in the lower layer is used as a reference picture, based on the discardable flag; and saving the picture in the lower layer in a decoded picture buffer, when the picture in the lower layer is used as the reference picture.

Description

Method and apparatus for encoding / decoding scalable video signal

The present invention relates to a method and apparatus for scalable video signal encoding / decoding.

Recently, the demand for high resolution and high quality images such as high definition (HD) and ultra high definition (UHD) images is increasing in various applications. As the video data becomes higher resolution and higher quality, the amount of data increases relative to the existing video data. Therefore, when the video data is transmitted or stored using a medium such as a conventional wired / wireless broadband line, The storage cost will increase. High efficiency image compression techniques can be used to solve these problems caused by high resolution and high quality image data.

An inter-screen prediction technique for predicting pixel values included in the current picture from a picture before or after the current picture using an image compression technique, an intra prediction technique for predicting pixel values included in a current picture using pixel information in the current picture, Various techniques exist, such as an entropy encoding technique for allocating a short code to a high frequency of appearance and a long code to a low frequency of appearance, and the image data can be effectively compressed and transmitted or stored.

Meanwhile, as the demand for high resolution video increases, the demand for stereoscopic video content also increases as a new video service. There is a discussion about a video compression technology for effectively providing high resolution and ultra high resolution stereoscopic image contents.

An object of the present invention is to provide a method and apparatus in which a picture of a lower layer is used as an inter-layer reference picture of a current picture of a higher layer in encoding / decoding a scalable video signal.

An object of the present invention is to provide a method and apparatus for upsampling a picture of a lower layer in encoding / decoding a scalable video signal.

An object of the present invention is to provide a method and apparatus for constructing a reference picture list using an interlayer reference picture in encoding / decoding a scalable video signal.

An object of the present invention is to provide a method and apparatus for effectively deriving texture information of a higher layer through inter-layer prediction in encoding / decoding a scalable video signal.

An object of the present invention is to efficiently manage a picture buffer decoded in a multilayer structure in encoding / decoding a scalable video signal.

The scalable video signal decoding method and apparatus according to the present invention obtain a discardable flag for a picture of a lower layer, and determine whether a picture of the lower layer is used as a reference picture based on the discardable flag. When the picture of the lower layer is used as a reference picture, the picture of the lower layer is stored in a decoded picture buffer.

The discardable flag according to the present invention is characterized in that it is information indicating whether a decoded picture is used as the reference picture in the process of decoding a picture having a lower priority in decoding order.

The discardable flag according to the present invention may be obtained from a slice segment header.

The discardable flag according to the present invention is obtained when the temporal level identifier of the picture of the lower layer is equal to or smaller than the maximum temporal level identifier for the lower layer.

The stored lower layer picture according to the present invention is marked as a short-term reference picture.

The scalable video signal encoding method and apparatus according to the present invention obtain a discardable flag for a picture of a lower layer, and determine whether the picture of the lower layer is used as a reference picture based on the discardable flag. When the picture of the lower layer is used as a reference picture, the picture of the lower layer is stored in a decoded picture buffer.

According to the present invention, the memory can be effectively managed by adaptively using the picture of the lower layer as the inter-layer reference picture of the current picture of the upper layer.

According to the present invention, it is possible to effectively upsample the picture of the lower layer.

According to the present invention, a reference picture list can be effectively constructed using an interlayer reference picture.

According to the present invention, texture information of an upper layer can be effectively derived through inter-layer prediction.

According to the present invention, the decoded picture buffer can be efficiently managed by storing the reference picture in the decoded picture buffer based on the discardable flag in the multilayer structure.

1 is a block diagram schematically illustrating an encoding apparatus according to an embodiment of the present invention.

2 is a block diagram schematically illustrating a decoding apparatus according to an embodiment of the present invention.

3 is a flowchart illustrating a process of performing inter-layer prediction of an upper layer using a corresponding picture of a lower layer according to an embodiment to which the present invention is applied.

4 is a diagram illustrating a process of determining whether a corresponding picture of a lower layer is used as an interlayer reference picture of a current picture as an embodiment to which the present invention is applied.

5 is a flowchart illustrating a method of upsampling a corresponding picture of a lower layer according to an embodiment to which the present invention is applied.

6 illustrates a method of extracting and obtaining a maximum temporal level identifier from a bitstream as an embodiment to which the present invention is applied.

FIG. 7 illustrates a method of deriving a maximum temporal level identifier for a lower layer using the maximum temporal level identifier for a previous layer as an embodiment to which the present invention is applied.

8 illustrates a method of deriving a maximum temporal level identifier based on a default temporal flag as an embodiment to which the present invention is applied.

FIG. 9 illustrates a method of managing a decoded picture buffer based on a discardable flag according to an embodiment to which the present invention is applied.

FIG. 10 illustrates a method of obtaining a discardable flag from a slice segment header according to an embodiment to which the present invention is applied.

FIG. 11 illustrates a method of obtaining a discardable flag based on a time level identifier according to an embodiment to which the present invention is applied.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Prior to this, terms or words used in the specification and claims should not be construed as having a conventional or dictionary meaning, and the inventors should properly explain the concept of terms in order to best explain their own invention. Based on the principle that can be defined, it should be interpreted as meaning and concept corresponding to the technical idea of the present invention. Therefore, the embodiments described in the specification and the drawings shown in the drawings are only the most preferred embodiment of the present invention and do not represent all of the technical idea of the present invention, various modifications that can be replaced at the time of the present application It should be understood that there may be equivalents and variations.

When a component is referred to herein as being “connected” or “connected” to another component, it may mean that it is directly connected to or connected to that other component, or another component in between. It may also mean that an element exists. In addition, the description "includes" a specific configuration in this specification does not exclude a configuration other than the configuration, it means that additional configuration may be included in the scope of the technical spirit of the present invention or the present invention.

Terms such as first and second may be used to describe various configurations, but the configurations are not limited by the terms. The terms are used to distinguish one configuration from another. For example, without departing from the scope of the present invention, the first configuration may be referred to as the second configuration, and similarly, the second configuration may also be referred to as the first configuration.

In addition, the components shown in the embodiments of the present invention are independently shown to represent different characteristic functions, and do not mean that each component is made of separate hardware or one software component unit. In other words, each component is listed as a component for convenience of description, and at least two of the components may form one component, or one component may be divided into a plurality of components to perform a function. The integrated and separated embodiments of each component are also included in the scope of the present invention without departing from the spirit of the present invention.

In addition, some of the components may not be essential components for performing essential functions in the present invention, but may be optional components for improving performance. The present invention can be implemented including only the components essential for implementing the essentials of the present invention except for the components used for improving performance, and the structure including only the essential components except for the optional components used for improving performance. Also included in the scope of the present invention.

Encoding and decoding of video that supports multiple layers in a bitstream is called scalable video coding. Since there is a strong correlation between the plurality of layers, the prediction may be performed by using this correlation to remove redundant elements of data and to improve encoding performance of an image. Performing prediction of the current layer using information of another layer is referred to as inter-layer prediction or inter-layer prediction in the following.

The plurality of layers may have different resolutions, where the resolution may mean at least one of spatial resolution, temporal resolution, and image quality. Resampling such as up-sampling or downsampling of a layer may be performed to adjust the resolution during inter-layer prediction.

The encoding apparatus 100 according to the present invention includes an encoder 100a for an upper layer and an encoder 100b for a lower layer.

The upper layer may be expressed as a current layer or an enhancement layer, and the lower layer may be expressed as an enhancement layer, a base layer, or a reference layer having a lower resolution than the upper layer. . The upper layer and the lower layer may have at least one of a spatial resolution, a temporal resolution according to a frame rate, and an image quality according to a color format or a quantization size. When a resolution change is necessary to perform inter-layer prediction, upsampling or downsampling of a layer may be performed.

The encoder 100a of the upper layer may include a divider 110, a predictor 120, a transformer 130, a quantizer 140, a reorderer 150, an entropy encoder 160, and an inverse quantizer ( 170, an inverse transform unit 180, a filter unit 190, and a memory 195.

The encoder 100b of the lower layer includes a divider 111, a predictor 125, a transformer 131, a quantizer 141, a reordering unit 151, an entropy encoder 161, and an inverse quantizer ( 171, an inverse transform unit 181, a filter unit 191, and a memory 196.

The encoder may be implemented by the image encoding method described in the following embodiments of the present invention, but operations in some components may not be performed to reduce the complexity of the encoding apparatus or for fast real time encoding. For example, in performing the intra prediction in the prediction unit, some limited number of methods are used without selecting the optimal intra intra coding method using all intra prediction modes in order to perform encoding in real time. A method of selecting one intra prediction mode among them as a final intra prediction mode using the intra prediction mode of the image may be used. As another example, it is possible to limit the use of the type of prediction block used in performing intra prediction or inter prediction.

The unit of a block processed by the encoding apparatus may be a coding unit that performs encoding, a prediction unit that performs prediction, or a transformation unit that performs transformation. A coding unit may be represented by a term such as a coding unit (CU), a prediction unit is a prediction unit (PU), and a transformation unit is a transform unit (TU).

The splitters 110 and 111 divide a layer image into a combination of a plurality of coding blocks, prediction blocks, and transform blocks, and one of the coding blocks, prediction blocks, and transform blocks according to a predetermined criterion (for example, a cost function). You can split the layer by selecting the combination of. For example, to split a coding unit in a layer image, a recursive tree structure such as a quad tree structure may be used. Hereinafter, in the embodiment of the present invention, the meaning of the coding block may be used not only as a block for encoding but also as a block for decoding.

The prediction block may be a unit for performing prediction such as intra prediction or inter prediction. The block for performing intra prediction may be a block having a square shape such as 2N × 2N or N × N. As a block for performing inter prediction, there is a prediction block partitioning method using Asymmetric Motion Partitioning (AMP), which is a square form such as 2Nx2N and NxN, or a rectangular form or asymmetric form such as 2NxN and Nx2N. Depending on the shape of the prediction block, the transform unit 115 may change a method of performing the transform.

The

prediction units

120 and 125 of the encoders 100a and 100b may include the intra prediction units 121 and 126 performing intra prediction and the inter prediction unit performing inter prediction. (122, 127). The predictor 120 of the higher layer encoder 100a may further include an inter-layer predictor 123 that performs prediction on the higher layer by using information of the lower layer.

The

prediction units

120 and 125 may determine whether to use inter prediction or intra prediction on the prediction block. In performing the intra prediction, the process of determining the intra prediction mode in units of prediction blocks and performing the intra prediction based on the determined intra prediction mode may be performed in units of transform blocks. The residual value (residual block) between the generated prediction block and the original block may be input to the

transformers

130 and 131. In addition, prediction mode information and motion information used for prediction may be encoded by the entropy encoder 130 together with the residual value and transmitted to the decoding apparatus.

When using the Pulse Coded Modulation (PCM) coding mode, the original block may be encoded as it is and transmitted to the decoder without performing prediction through the

prediction units

120 and 125.

The intra prediction units 121 and 126 may generate an intra prediction block based on reference pixels present around the current block (the block to be predicted). In the intra prediction method, the intra prediction mode may have a directional prediction mode using a reference pixel according to a prediction direction and a non-directional mode without considering the prediction direction. The mode for predicting luma information and the mode for predicting color difference information may be different. In order to predict the color difference information, an intra prediction mode in which luma information is predicted or predicted luma information may be used. If the reference pixel is not available, the unusable reference pixel may be replaced with another pixel, and a prediction block may be generated using the reference pixel.

The prediction block may include a plurality of transform blocks. If the prediction block has the same size as the transform block when the intra prediction is performed, pixels present on the left side of the prediction block, pixels present on the upper left side, and top Intra-prediction of the prediction block may be performed based on the pixels present in the. However, when the prediction block is different from the size of the transform block when the intra prediction is performed, and a plurality of transform blocks are included in the prediction block, intra prediction is performed by using neighboring pixels adjacent to the transform block as reference pixels. Can be done. Here, the neighboring pixel adjacent to the transform block may include at least one of the neighboring pixel adjacent to the prediction block and the pixels already decoded in the prediction block.

The intra prediction method may generate a prediction block after applying a mode dependent intra smoothing (MDIS) filter to a reference pixel according to the intra prediction mode. The type of MDIS filter applied to the reference pixel may be different. The MDIS filter is an additional filter applied to the predicted block in the picture by performing the intra prediction and may be used to reduce the residual present in the predicted block in the picture generated after performing the prediction with the reference pixel. In performing MDIS filtering, filtering on a reference pixel and some columns included in the predicted block in the screen may perform different filtering according to the direction of the intra prediction mode.

The

inter prediction units

122 and 127 may perform prediction by referring to information of a block included in at least one of a previous picture or a subsequent picture of the current picture. The

inter prediction units

122 and 127 may include a reference picture interpolator, a motion predictor, and a motion compensator.

The reference picture interpolation unit may receive reference picture information from the

memories

195 and 196 and generate pixel information of an integer pixel or less in the reference picture. In the case of a luma pixel, a DCT-based 8-tap interpolation filter having different filter coefficients may be used to generate pixel information of integer pixels or less in units of 1/4 pixels. In the case of a chrominance signal, a DCT-based interpolation filter having different filter coefficients may be used to generate pixel information of an integer pixel or less in units of 1/8 pixels.

The

inter prediction units

122 and 127 may perform motion prediction based on the reference picture interpolated by the reference picture interpolator. As a method for calculating a motion vector, various methods such as a full search-based block matching algorithm (FBMA), a three step search (TSS), and a new three-step search algorithm (NTS) may be used. The motion vector may have a motion vector value of 1/2 or 1/4 pixel units based on the interpolated pixels. The

inter prediction units

122 and 127 may perform prediction on the current block by applying one inter prediction method among various inter prediction methods.

For example, various methods, such as a skip method, a merge method, and a motion vector predictor (MVP), may be used as the inter prediction method.

In inter prediction, motion information, that is, information such as a reference index, a motion vector, and a residual signal, is entropy coded and transmitted to a decoder. When the skip mode is applied, since the residual signal is not generated, the conversion and quantization processes for the residual signal may be omitted.

The interlayer prediction unit 123 performs interlayer prediction for predicting an upper layer by using information of a lower layer. The inter-layer prediction unit 123 may perform inter-layer prediction using texture information, motion information, etc. of the lower layer.

In inter-layer prediction, prediction of a current block of an upper layer may be performed using motion information on a picture of a lower layer (reference layer) using a picture of a lower layer as a reference picture. The picture of the reference layer used as the reference picture in inter-layer prediction may be a picture sampled according to the resolution of the current layer. In addition, the motion information may include a motion vector and a reference index. In this case, the value of the motion vector for the picture of the reference layer may be set to (0,0).

As an example of inter-layer prediction, a prediction method using a picture of a lower layer as a reference picture has been described, but the present invention is not limited thereto. The inter-layer prediction unit 123 may perform inter-layer texture prediction, inter-layer motion prediction, inter-layer syntax prediction, and inter-layer difference prediction.

Inter-layer texture prediction may derive the texture of the current layer based on the texture of the reference layer. The texture of the reference layer may be sampled according to the resolution of the current layer, and the inter-layer predictor 123 may predict the texture of the current layer based on the sampled texture of the reference layer.

Inter-layer motion prediction may derive the motion vector of the current layer based on the motion vector of the reference layer. In this case, the motion vector of the reference layer may be scaled according to the resolution of the current layer. In inter-layer syntax prediction, the syntax of the current layer may be predicted based on the syntax of the reference layer. For example, the inter-layer prediction unit 123 may use the syntax of the reference layer as the syntax of the current layer. In addition, in the difference prediction between layers, the picture of the current layer may be reconstructed using the difference between the reconstructed image of the reference layer and the reconstructed image of the current layer.

A residual block including residual information, which is a difference between the predicted block generated by the

predictors

120 and 125 and the reconstructed block of the predicted block, is generated, and the residual block is input to the

transformers

130 and 131.

The

transform units

130 and 131 may transform the residual block using a transform method such as a discrete cosine transform (DCT) or a discrete sine transform (DST). Whether DCT or DST is applied to transform the residual block may be determined based on intra prediction mode information of the prediction block used to generate the residual block and size information of the prediction block. That is, the

transformers

130 and 131 may apply the transformation method differently according to the size of the prediction block and the prediction method.

The

quantizers

140 and 141 may quantize the values transformed by the

transformers

130 and 131 into the frequency domain. The quantization coefficient may change depending on the block or the importance of the image. The values calculated by the

quantizers

140 and 141 may be provided to the dequantizers 170 and 17 and the

reordering units

150 and 151.

The

reordering units

150 and 151 may reorder coefficient values with respect to the quantized residual value. The

reordering units

150 and 151 may change the two-dimensional block shape coefficients into a one-dimensional vector form through a coefficient scanning method. For example, the

realignment units

150 and 151 may scan DC coefficients to coefficients in the high frequency region by using a Zig-Zag scan method and change them into one-dimensional vectors. Depending on the size of the transform block and the intra prediction mode, a vertical scan method for scanning two-dimensional block shape coefficients in a column direction, not a zig-zag scan method, and a horizontal scan method for scanning two-dimensional block shape coefficients in a row direction Can be used. That is, according to the size of the transform block and the intra prediction mode, it is possible to determine which scan method among zigzag-scan, vertical scan and horizontal scan is used.

The

entropy encoders

160 and 161 may perform entropy encoding based on the values calculated by the

reordering units

150 and 151. Entropy encoding may use various encoding methods such as, for example, Exponential Golomb, Context-Adaptive Variable Length Coding (CAVLC), and Context-Adaptive Binary Arithmetic Coding (CABAC).

The

entropy encoders

160 and 161 transmit residual value coefficient information, block type information, prediction mode information, partition unit information, prediction block information, and the like of the coding block from the

reordering units

150 and 151 and the

prediction units

120 and 125. Entropy encoding may be performed based on a predetermined encoding method by receiving various information such as unit information, motion information, reference frame information, block interpolation information, and filtering information. In addition, the

entropy encoder

160 or 161 may entropy-encode coefficient values of coding units input from the

reordering unit

150 or 151.

The

entropy encoders

160 and 161 may encode the intra prediction mode information of the current block by performing binarization on the intra prediction mode information. The

entropy encoder

160 or 161 may include a codeword mapping unit for performing such a binarization operation, and may perform different binarization according to the size of a prediction block for performing intra prediction. In the codeword mapping unit, the codeword mapping table may be adaptively generated or stored in advance through a binarization operation. As another embodiment, the

entropy encoders

160 and 161 may express prediction mode information in the current screen using a codenum mapping unit for performing codenum mapping and a codeword mapping unit for performing codeword mapping. In the codenum mapping unit and the codeword mapping unit, a codenum mapping table and a codeword mapping table may be generated or stored.

The

inverse quantizers

170 and 171 and the

inverse transformers

180 and 181 inverse quantize the quantized values in the

quantizers

140 and 141 and inversely transform the converted values in the

transformers

130 and 131. The residual values generated by the

inverse quantizers

170 and 171 and the

inverse transformers

180 and 181 may be predicted by the motion estimator, the motion compensator, and the intra prediction unit included in the

predictors

120 and 125. It may be combined with the prediction block to generate a reconstructed block.

The

filter units

190 and 191 may include at least one of a deblocking filter and an offset correction unit.

The deblocking filter may remove block distortion caused by boundaries between blocks in the reconstructed picture. In order to determine whether to perform deblocking, it may be determined whether to apply a deblocking filter to the current block based on the pixels included in several columns or rows included in the block. When the deblocking filter is applied to the block, a strong filter or a weak filter may be applied according to the required deblocking filtering strength. In addition, in applying the deblocking filter, horizontal filtering and vertical filtering may be performed in parallel when vertical filtering and horizontal filtering are performed.

The offset correction unit may correct the offset with respect to the original image on a pixel-by-pixel basis for the deblocking image. In order to perform offset correction for a specific picture, the pixels included in the image are divided into predetermined areas, and then, the area to be offset is determined and the offset is applied to the corresponding area, or the offset is applied considering the edge information of each pixel. Can be used.

The

filter units

190 and 191 may apply only the deblocking filter or both the deblocking filter and the offset correction without applying both the deblocking filter and the offset correction.

The

memories

195 and 196 may store reconstructed blocks or pictures calculated by the

filters

190 and 191, and the stored reconstructed blocks or pictures may be provided to the

predictors

120 and 125 when performing inter prediction. have.

The information output from the entropy encoder 100b of the lower layer and the information output from the entropy encoder 100a of the upper layer may be multiplexed by the MUX 197 and output as a bitstream.

The MUX 197 may be included in the encoder 100a of the upper layer or the encoder 100b of the lower layer, or may be implemented as an independent device or module separate from the encoder 100.

As shown in FIG. 2, the decoding apparatus 200 includes a decoder 200a of an upper layer and a decoder 200b of a lower layer.

The decoder 200a of the upper layer includes an entropy decoder 210, a reordering unit 220, an inverse quantization unit 230, an inverse transform unit 240, a prediction unit 250, a filter unit 260, and a memory 270. ) May be included.

The lower layer decoding unit 200b includes an entropy decoding unit 211, a reordering unit 221, an inverse quantization unit 231, an inverse transform unit 241, a prediction unit 251, a filter unit 261, and a memory 271. ) May be included.

When a bitstream including a plurality of layers is transmitted from the encoding apparatus, the DEMUX 280 may demultiplex information for each layer and transmit the information to the

decoders

200a and 200b for each layer. The input bitstream may be decoded in a procedure opposite to that of the encoding apparatus.

The

entropy decoders

210 and 211 may perform entropy decoding in a procedure opposite to that of the entropy encoder in the encoding apparatus. Information for generating a prediction block among the information decoded by the

entropy decoders

210 and 211 is provided to the

predictors

250 and 251, and the residual value obtained by entropy decoding by the

entropy decoders

210 and 211 is a reordering unit. It may be input to (220, 221).

Like the

entropy encoders

160 and 161, the

entropy decoders

210 and 211 may use at least one of CABAC and CAVLC.

The

entropy decoders

210 and 211 may decode information related to intra prediction and inter prediction performed by the encoding apparatus. The

entropy decoder

210 or 211 may include a codeword mapping unit and include a codeword mapping table for generating a received codeword as an intra prediction mode number. The codeword mapping table may be stored in advance or generated adaptively. When using the codenum mapping table, a codenum mapping unit for performing codenum mapping may be additionally provided.

The

reordering units

220 and 221 may reorder the bitstreams entropy decoded by the

entropy decoding units

210 and 211 based on a method of rearranging the bitstreams by the encoder. Coefficients expressed in the form of a one-dimensional vector may be reconstructed by reconstructing the coefficients in a two-dimensional block form. The

reordering units

220 and 221 may be realigned by receiving information related to coefficient scanning performed by the encoder and performing reverse scanning based on the scanning order performed by the corresponding encoder.

The

inverse quantization units

230 and 231 may perform inverse quantization based on quantization parameters provided by the encoding apparatus and coefficient values of the rearranged block.

The

inverse transformers

240 and 241 may perform inverse DCT or inverse DST on the DCT or DST performed by the

transformers

130 and 131 with respect to the quantization result performed by the encoding apparatus. The inverse transform may be performed based on a transmission unit determined by the encoding apparatus. The DCT and the DST may be selectively performed by the transform unit of the encoding apparatus according to a plurality of pieces of information, such as a prediction method, a size of the current block, and a prediction direction. The

inverse transformers

240 and 241 of the decoding apparatus may convert Inverse transformation may be performed based on the performed transformation information. When the transform is performed, the transform may be performed based on the coding block rather than the transform block.

The

prediction units

250 and 251 may generate the prediction blocks based on the prediction block generation related information provided by the

entropy decoding units

210 and 211 and previously decoded blocks or picture information provided by the

memories

270 and 271. .

The

predictors

250 and 251 may include a prediction unit determiner, an inter prediction unit, and an intra prediction unit.

The prediction unit discriminator receives various information such as prediction unit information input from the entropy decoder, prediction mode information of the intra prediction method, and motion prediction related information of the inter prediction method, and distinguishes the prediction block from the current coding block. It is possible to determine whether to perform this inter prediction or intra prediction.

The inter prediction unit uses information required for inter prediction of the current prediction block provided by the encoding apparatus to the current prediction block based on information included in at least one of a previous picture or a subsequent picture of the current picture including the current prediction block. Inter prediction can be performed. In order to perform inter prediction, a motion prediction method of a prediction block included in a coding block based on a coding block uses a skip mode, a merge mode, a motion vector predictor (MVP) (AMVP). Mode) can be determined.

The intra prediction unit may generate a prediction block based on the reconstructed pixel information in the current picture. When the prediction block is a prediction block that performs intra prediction, intra prediction may be performed based on intra prediction mode information of the prediction block provided by the encoding apparatus. The intra prediction unit is an MDIS filter that performs filtering on the reference pixel of the current block, a reference pixel interpolator that generates a reference pixel of an integer value or less by interpolating the reference pixel, and filters when the prediction mode of the current block is DC mode. It may include a DC filter for generating a prediction block through.

The predictor 250 of the upper layer decoder 200a may further include an inter-layer predictor that performs inter-layer prediction for predicting an upper layer by using information of the lower layer.

The inter-layer prediction unit may perform inter-layer prediction using intra prediction mode information and motion information.

In inter-layer prediction, prediction of a current block of an upper layer may be performed using motion information of a lower layer (reference layer) picture using a picture of a lower layer as a reference picture.

The picture of the reference layer used as the reference picture in inter-layer prediction may be a picture sampled according to the resolution of the current layer. In addition, the motion information may include a motion vector and a reference index. In this case, the value of the motion vector for the picture of the reference layer may be set to (0,0).

As an example of inter-layer prediction, a prediction method using a picture of a lower layer as a reference picture has been described, but the present invention is not limited thereto. The inter-layer prediction unit 123 may further perform inter-layer texture prediction, inter-layer motion prediction, inter-layer syntax prediction, and inter-layer difference prediction.

Inter-layer texture prediction may derive the texture of the current layer based on the texture of the reference layer. The texture of the reference layer may be sampled according to the resolution of the current layer, and the inter-layer predictor may predict the texture of the current layer based on the sampled texture. Inter-layer motion prediction may derive the motion vector of the current layer based on the motion vector of the reference layer. In this case, the motion vector of the reference layer may be scaled according to the resolution of the current layer. In inter-layer syntax prediction, the syntax of the current layer may be predicted based on the syntax of the reference layer. For example, the inter-layer prediction unit 123 may use the syntax of the reference layer as the syntax of the current layer. In addition, in the difference prediction between layers, the picture of the current layer may be reconstructed using the difference between the reconstructed image of the reference layer and the reconstructed image of the current layer.

The reconstructed block or picture may be provided to the

filter units

260 and 261. The

filter units

260 and 261 may include a deblocking filter and an offset correction unit.

Information about whether a deblocking filter is applied to the corresponding block or picture and when the deblocking filter is applied to the block or picture may be provided from the encoding apparatus as to whether a strong filter or a weak filter is applied. The deblocking filter of the decoding apparatus may receive the deblocking filter related information provided by the encoding apparatus and perform the deblocking filtering on the corresponding block in the decoding apparatus.

The offset correction unit may perform offset correction on the reconstructed image based on the type of offset correction and offset value information applied to the image during encoding.

The

memories

270 and 271 may store the reconstructed picture or block to be used as the reference picture or the reference block, and output the reconstructed picture.

The encoding apparatus and the decoding apparatus may encode three or more layers instead of two layers. In this case, a plurality of encoders for a higher layer and a decoder for a higher layer may be provided in correspondence to the number of upper layers. Can be.

In SVC (Scalable Video Coding) supporting multi-layer structure, there is an association between layers. Prediction using this association can remove redundant elements of data and improve the encoding performance of the image.

Therefore, when predicting a picture (image) of a current layer (enhanced layer) to be encoded / decoded, not only inter prediction or intra prediction using information of the current layer, but also inter layer prediction using information of another layer may be performed. .

When inter-layer prediction is performed, the current layer may generate a prediction sample of the current layer by using a decoded picture of a reference layer used for inter-layer prediction as a reference picture.

In this case, since the current layer and the reference layer may have at least one of spatial resolution, temporal resolution, and image quality different from each other (that is, due to scalability differences between layers), the picture of the decoded reference layer matches the scalability of the current layer. Resampling may be performed and then used as a reference picture for inter-layer prediction of the current layer. Resampling means up-sampling or downsampling samples of a reference layer picture according to a picture size of a current layer.

In the present specification, the current layer refers to a layer on which current encoding or decoding is performed, and may be an enhancement layer or an upper layer. The reference layer refers to a layer referenced by the current layer for inter-layer prediction and may be a base layer or a lower layer. A picture (ie, a reference picture) of a reference layer used for inter layer prediction of the current layer may be referred to as an inter layer reference picture or an inter layer reference picture.

Referring to FIG. 3, it may be determined whether a corresponding picture of a lower layer is used as an inter-layer reference picture for the current picture of the upper layer based on the temporal identifier TemporalID of the lower layer (S300).

For example, if the temporal resolution of the current picture that you want to encode in the enhancement layer is low (that is, if the temporal ID (TemporalID) of the current picture has a small value), the other picture and display already decoded in the enhancement layer are displayed. The order difference becomes large. In this case, it is more likely that the picture characteristics between the current picture and the already decoded picture are different. Therefore, rather than using the already decoded pictures of the enhancement layer as the reference picture, the upsampled picture in the lower layer is used as the reference picture. The chances are high.

On the other hand, when the temporal resolution of the current picture to be encoded in the enhancement layer is high (that is, when the temporal ID (TemporalID) of the current picture has a large value), the display order difference is different from other pictures already decoded in the enhancement layer. Is not large. In this case, the image characteristics between the current picture and the already decoded picture are more likely to be similar. Therefore, rather than using the upsampled picture in the lower layer as the reference picture, the already decoded pictures in the enhancement layer are used as the reference picture. The chances are high.

As such, when the temporal resolution of the current picture is low, the inter-layer inter prediction method is effective. Therefore, it is necessary to determine whether to allow inter-layer inter prediction in consideration of a specific temporal identifier (TemporalID) of the lower layer. To this end, it is possible to signal the maximum temporal level identifier of the lower layer that allows inter-layer prediction. This will be described in detail with reference to FIG. 4.

The corresponding picture of the lower layer may mean a picture located in the same time zone as the current picture of the upper layer. For example, the corresponding picture may refer to a picture having the same picture order count (POC) information as the current picture of the upper layer. The corresponding picture of the lower layer may be included in the same Access Unit (AU) as the current picture of the upper layer.

In addition, the video sequence may include a plurality of layers that are coded scalable according to temporal / spatial resolution or quantization size. The temporal level identifier may mean an identifier that specifies each of a plurality of layers that are coded scalable according to temporal resolution. Accordingly, the plurality of layers included in the video sequence may have the same temporal identifier or may have different temporal identifiers, respectively.

In operation S310, a reference picture list of the current picture may be generated according to the determination in operation S300.

In detail, when it is determined that the corresponding picture of the lower layer is used as the interlayer reference picture of the current picture, the interlayer reference picture may be generated by upsampling the corresponding picture. A process of upsampling the corresponding picture of the lower layer will be described in detail with reference to FIG. 5.

A reference picture list including the generated interlayer reference picture may be generated. For example, a reference picture list may be constructed using a reference picture belonging to the same layer as the current block, that is, a temporal reference picture, and an interlayer reference picture may be arranged after the temporal reference picture.

Alternatively, an interlayer reference picture may be added between temporal reference pictures. For example, the interlayer reference picture may be arranged after the first temporal reference picture in the reference picture list composed of the temporal reference picture. The first temporal reference picture in the reference picture list may mean a reference picture having a reference index of zero. As such, when the interlayer reference picture is arranged after the first temporal reference picture, the temporal reference pictures except the first temporal reference picture may be arranged after the interlayer reference picture.

On the other hand, if it is determined that the corresponding picture of the lower layer is not used as the interlayer reference picture of the current picture, the corresponding picture is not included in the reference picture list of the current picture. That is, the reference picture list of the current picture may be composed of a reference picture belonging to the same layer as the current picture, that is, a temporal reference picture. As described above, since the picture of the lower layer can be excluded from the decoded picture buffer (DPB), the decoded picture buffer can be managed efficiently.

Inter-prediction may be performed on the current block based on the reference picture list generated in step S310 (S320).

In detail, the reference picture may be specified in the generated reference picture list using the reference index of the current block. In addition, a reference block within a reference picture may be specified using the motion vector of the current block. The current block may perform inter prediction using the specified reference block.

Alternatively, when the current block uses the interlayer reference picture as the reference picture, the current block may perform inter-layer prediction using blocks of the same position in the interlayer reference picture. To this end, when the reference index of the current block specifies the interlayer reference picture in the reference picture list, the motion vector of the current block may be set to (0,0).

Referring to FIG. 4, a maximum temporal level identifier for a lower layer may be obtained (S400).

Here, the maximum temporal level identifier may mean the maximum value of the temporal level identifier of the lower layer that allows inter-layer prediction of the upper layer.

The maximum temporal level identifier may be obtained by extracting directly from the bitstream. Alternatively, it may be derived using the maximum temporal identifier of the previous layer. Alternatively, the information may be obtained based on a predefined default time level value. Alternatively, it may be obtained based on a default time level flag. A detailed method of obtaining the maximum time level identifier will be described with reference to FIGS. 6 to 8.

The maximum temporal level identifier obtained in step S400 and the temporal level identifiers of the lower layers may be compared to determine whether the corresponding picture of the lower layer is used as an interlayer reference picture of the current picture (S410).

For example, when the temporal level identifier of the lower layer is larger than the maximum temporal level identifier, the corresponding picture of the lower layer may not be used as an interlayer reference picture of the current picture. That is, the current picture does not perform inter-layer prediction by using the corresponding picture of the lower layer.

On the other hand, when the temporal level identifier of the lower layer is less than or equal to the maximum temporal level identifier, the corresponding picture of the lower layer may be used as an interlayer reference picture of the current picture. That is, the current picture may perform inter-layer prediction using a picture of a lower layer having a temporal level identifier smaller than the maximum temporal level identifier.

Referring to FIG. 5, the reference sample position of the lower layer corresponding to the current sample position of the upper layer may be derived (S500).

Since the resolutions of the upper layer and the lower layer may be different, a reference sample position corresponding to the current sample position may be derived in consideration of the resolution difference between the two layers. That is, the aspect ratio may be considered between the picture of the upper layer and the picture of the lower layer. In addition, since the upsampled picture of the lower layer may not match the size of the picture of the upper layer, an offset may be required to correct this.

For example, the reference sample position may be derived in consideration of the scale factor and the upsampled lower layer offset.

Here, the scale factor may be calculated based on a ratio of the width and the height between the current picture of the upper layer and the corresponding picture of the lower layer.

The upsampled lower layer offset may mean position difference information between any one sample located at the edge of the current picture and any one sample located at the edge of the interlayer reference picture. For example, the upsampled lower layer offset includes horizontal position information in the horizontal / vertical direction between the upper left sample of the current picture and the upper left sample of the interlayer reference picture, and the lower right sample of the current picture and the lower right sample of the interlayer reference picture. Position difference information in the horizontal / vertical direction of the liver may be included.

The upsampled lower layer offset may be obtained from the bitstream. For example, the upsampled lower layer offset may be obtained from at least one of a video parameter set, a sequence parameter set, a picture parameter set, and a slice header. Can be.

The filter coefficient of the upsampling filter may be determined in consideration of the phase of the reference sample position derived in step S500 (S510).

Here, the upsampling filter may be any one of a fixed upsampling filter and an adaptive upsampling filter.

1. Fixed Upsampling Filter

The fixed upsampling filter may mean an upsampling filter having a predetermined filter coefficient without considering the feature of the image. A tap filter may be used as the fixed upsampling filter, which may be defined for the luminance component and the chrominance component, respectively. Hereinafter, a fixed upsampling filter having an accuracy of 1/16 sample units will be described with reference to Tables 1 to 2.

Table 1

Phase p	Interpolation filter coefficients
Phase p	f [p, 0]	f [p, 1]	f [p, 2]	f [p, 3]	f [p, 4]	f [p, 5]	f [p, 6]	f [p, 7]
0	0	0	0	64	0	0	0	0
One	0	One	-3	63	4	-2	One	0
2	-One	2	-5	62	8	-3	One	0
3	-One	3	-8	60	13	-4	One	0
4	-One	4	-10	58	17	-5	One	0
5	-One	4	-11	52	26	-8	3	-One
6	-One	3	-3	47	31	-10	4	-One
7	-One	4	-11	45	34	-10	4	-One
8	-One	4	-11	40	40	-11	4	-One
9	-One	4	-10	34	45	-11	4	-One
10	-One	4	-10	31	47	-9	3	-One
11	-One	3	-8	26	52	-11	4	-One
12	0	One	-5	17	58	-10	4	-One
13	0	One	-4	13	60	-8	3	-One
14	0	One	-3	8	62	-5	2	-One
15	0	One	-2	4	63	-3	One	0

Table 1 is a table that defines the filter coefficients of the fixed upsampling filter for the luminance component.

As shown in Table 1, in the case of upsampling the luminance component, an 8-tap filter is applied. That is, interpolation may be performed using a reference sample of a reference layer corresponding to the current sample of the upper layer and a neighboring sample adjacent to the reference sample. Here, the neighbor sample may be specified according to the direction in which interpolation is performed. For example, when performing interpolation in the horizontal direction, the neighboring sample may include three consecutive samples to the left and four consecutive samples to the right based on the reference sample. Alternatively, when performing interpolation in the vertical direction, the neighboring sample may include three consecutive samples at the top and four consecutive samples at the bottom based on the reference sample.

In addition, since interpolation is performed with an accuracy of 1/16 samples, there are 16 phases in total. This is to support resolutions of various magnifications such as 2 times and 1.5 times.

In addition, the fixed upsampling filter may use different filter coefficients for each phase p. Except in the case where phase p is zero, the magnitude of each filter coefficient may be defined to fall in the range of 0 to 63. This means that the filtering is performed with a precision of 6 bits. Here, a phase p of 0 means a position of an integer multiple of n times when interpolated in units of 1 / n samples.

TABLE 2

Phase p	Interpolation filter coefficients
Phase p	f [p, 0]	f [p, 1]	f [p, 2]	f [p, 3]
0	0	64	0	0
One	-2	62	4	0
2	-2	58	10	-2
3	-4	56	14	-2
4	-4	54	16	-2
5	-6	52	20	-2
6	-6	46	28	-4
7	-4	42	30	-4
8	-4	36	36	-4
9	-4	30	42	-4
10	-4	28	46	-6
11	-2	20	52	-6
12	-2	16	54	-4
13	-2	14	56	-4
14	-2	10	58	-2
15	0	4	62	-2

Table 2 is a table that defines the filter coefficients of the fixed upsampling filter for the chrominance components.

As shown in Table 2, in the case of upsampling for the chrominance component, a 4-tap filter may be applied unlike the luminance component. That is, interpolation may be performed using a reference sample of a reference layer corresponding to the current sample of the upper layer and a neighboring sample adjacent to the reference sample. Here, the neighbor sample may be specified according to the direction in which interpolation is performed. For example, when performing interpolation in the horizontal direction, the neighboring sample may include one sample to the left and two samples to the right based on the reference sample. Alternatively, when performing interpolation in the vertical direction, the neighboring sample may include one sample continuous to the top and two samples continuous to the bottom based on the reference sample.

Meanwhile, since the interpolation is performed with the accuracy of 1/16 sample unit like the luminance component, a total of 16 phases exist, and different filter coefficients may be used for each phase p. And, except for the case where the phase p is 0, the magnitude of each filter coefficient may be defined to be in the range of 0 to 62. This also means filtering with 6bits precision.

Although an example in which an 8-tap filter is applied to the luminance component and a 4-tap filter to the chrominance component has been described as an example, the present invention is not limited thereto, and the order of the tap filter may be variably determined in consideration of coding efficiency. Of course.

2. Adaptive Upsampling Filter

Instead of using fixed filter coefficients, an optimal filter coefficient may be determined by an encoder in consideration of characteristics of an image, signaled, and transmitted to a decoder. In this way, the adaptive upsampling filter uses the filter coefficients that are adaptively determined in the encoder. Since the characteristics of the image are different in picture units, coding efficiency can be improved by using an adaptive upsampling filter that can express the characteristics of the image better than using a fixed upsampling filter in all cases.

The interlayer reference picture may be generated by applying the filter coefficient determined in operation S510 to the corresponding picture of the lower layer in operation S520.

Specifically, interpolation may be performed by applying the determined filter coefficients of the upsampling filter to samples of the corresponding picture. Here, the interpolation may be performed primarily in the horizontal direction, and may be performed in the vertical direction secondary to the sample generated after the horizontal interpolation.

The encoder may determine an optimal maximum temporal level identifier, encode it, and transmit the same to the decoder. In this case, the encoder may encode the determined maximum temporal level identifier as it is or may encode a value obtained by adding 1 to the determined maximum temporal level identifier (max_tid_il_ref_pics_plus1, hereinafter referred to as maximum temporal indicator).

Referring to FIG. 6, the maximum temporal indicator for the lower layer may be obtained from the bitstream (S600).

Here, the maximum temporal level indicator may be obtained as many as the maximum number of layers allowed in one video sequence. The maximum temporal indicator may be obtained from a video parameter set of the bitstream.

Specifically, when the value of the obtained maximum temporal indicator is 0, this may mean that the corresponding picture of the lower layer is not used as the interlayer reference picture of the upper layer. In this case, the corresponding picture of the lower layer may be a non-random access picture.

For example, when the value of the maximum temporal indicator is 0, the picture of the i-th layer among the plurality of layers of the video sequence is not used as a reference picture for inter-layer prediction of the picture belonging to the (i + 1) th layer.

On the other hand, when the value of the maximum temporal indicator is greater than 0, this may mean that a corresponding picture of a lower layer having a temporal identifier greater than the maximum temporal identifier is not used as an interlayer reference picture of the upper layer.

For example, if the value of the maximum temporal indicator is greater than 0, a picture belonging to the i th layer among the plurality of layers of the video sequence and having a temporal level identifier having a value greater than the maximum temporal identifier is (i + 1). It is not used as a reference picture for inter-layer prediction of a picture belonging to the first layer. In other words, if the value of the maximum temporal indicator is greater than 0 and the picture belonging to the i th layer among the plurality of layers of the video sequence has a temporal level identifier of a value smaller than the maximum temporal identifier, the (i + 1) th It may be used as a reference picture for inter-layer prediction of a picture belonging to a layer. Here, the maximum temporal level identifier is a value derived from the maximum temporal level indicator. For example, the maximum temporal level identifier may be derived by subtracting 1 from the value of the maximum temporal level indicator.

Meanwhile, the maximum time level indicator extracted in step S600 has a value (eg, 0 to 7) within a predetermined range. If the value of the maximum temporal indicator extracted in step S600 corresponds to a maximum value within a predetermined range, the corresponding picture of the lower layer is higher than the temporal ID (TemporalID) of the corresponding picture of the lower layer. It may be used as an interlayer reference picture of a layer.

Instead of encoding the maximum temporal level identifier (or the maximum temporal indicator) for the lower layer as it is, the maximum temporal level identifier (or , The amount of bits required to encode the maximum temporal indicator) can be reduced. Here, the previous layer may mean a layer having a lower resolution than the lower layer.

Referring to FIG. 7, the maximum temporal level governor max_tid_il_ref_pics_plus1 [0] for the lowest layer among the plurality of layers in the video sequence may be obtained (S700). This is because, for the lowest layer in the video sequence, there is no previous layer to reference to derive the maximum temporal identifier.

Here, if the value of the maximum temporal level governor (max_tid_il_ref_pics_plus1 [0]) is 0, the picture of the lowest layer (that is, the layer whose i = 0) in the video sequence is inter-layer prediction of the picture belonging to the (i + 1) th layer. It should not be used as a reference picture for.

On the other hand, if the value of the maximum temporal indicator (max_tid_il_ref_pics_plus1 [0]) is greater than 0, a picture belonging to the lowest layer in the video sequence and having a temporal level identifier having a value greater than the maximum temporal level identifier are (i + 1) th. It is not used as a reference picture for inter-layer prediction of pictures belonging to a layer. Therefore, when the value of the maximum temporal indicator max_tid_il_ref_pics_plus1 [0] is greater than 0 and a picture belonging to the lowest layer of the video sequence has a temporal level identifier of a value smaller than the maximum temporal level identifier (i + 1) It may be used as a reference picture for inter-layer prediction of the picture belonging to the first layer. Here, the maximum time level identifier is a value derived from the maximum time level indicator (max_tid_il_ref_pics_plus1 [0]). For example, the maximum time level identifier is derived by subtracting 1 from the value of the maximum time level indicator (max_tid_il_ref_pics_plus1 [0]). Can be.

The maximum time level indicator max_tid_il_ref_pics_plus1 [0] has a value (eg, 0 to 7) within a predetermined range. If the value of the maximum temporal indicator (max_tid_il_ref_pics_plus1 [0]) corresponds to the maximum value within a predetermined range, the corresponding picture of the lowest layer regardless of the temporal ID (TemporalID) of the corresponding picture of the lowest layer May be used as the inter-layer reference picture of the (i + 1) th layer.

Referring to FIG. 7, a differential time level indicator (delta_max_tid_il_ref_pics_plus1 [i]) for each of the remaining layers except the lowest layer in the video sequence may be obtained (S710).

Here, the difference time level indicator may mean a difference value between the maximum time level indicator (max_tid_il_ref_pics_plus1 [i]) for the i-th layer and the maximum time level indicator (max_tid_il_ref_pics_plus1 [i-1]) for the (i-1) th layer. Can be.

In this case, the maximum temporal indicator (max_tid_il_ref_pics_plus1 [i]) for the i-th layer is obtained from the obtained differential temporal indicator (delta_max_tid_il_ref_pics_plus1 [i]) and the maximum temporal indicator (max_tid_il_ref_pics_plus1 [i]) for the (i-1) th layer. -1]).

As shown in FIG. 6, when the value of the maximum temporal indicator (max_tid_il_ref_pics_plus1 [i]) for the derived i-th layer is 0, the picture of the i-th layer among the plurality of layers of the video sequence is (i + 1). It is not used as a reference picture for inter-layer prediction of a picture belonging to the first layer.

On the other hand, if the value of the maximum temporal indicator (max_tid_il_ref_pics_plus1 [i]) is greater than 0, a picture belonging to the i th layer among the plurality of layers of the video sequence and having a temporal level identifier having a value greater than the maximum temporal identifier is ( It is not used as a reference picture for inter-layer prediction of a picture belonging to the i + 1) th layer. (I +) only when the value of the maximum temporal indicator (max_tid_il_ref_pics_plus1 [i]) is greater than 0 and a picture belonging to the i th layer among the plurality of layers of the video sequence has a temporal level identifier having a value smaller than the maximum temporal level identifier. It may be used as a reference picture for inter-layer prediction of the picture belonging to the 1) th layer. Here, the maximum temporal level identifier is a value derived from the maximum temporal level indicator. For example, the maximum temporal level identifier may be derived by subtracting 1 from the value of the maximum temporal level indicator.

Meanwhile, the derived maximum temporal level indicator max_tid_il_ref_pics_plus1 [i] has a value within a predetermined range (for example, 0 to 7). When the value of the derived maximum temporal level indicator (max_tid_il_ref_pics_plus1 [i]) corresponds to the maximum value within the predetermined range, the correspondence of the i th layer regardless of the temporal ID (TemporalID) of the corresponding picture of the i th layer The picture may be used as an interlayer reference picture of the (i + 1) th layer.

The differential time level indicator extracted in step S710 may have a value within a predetermined range. In detail, when the frame rate difference between the i th layer and the (i-1) th layer is large, the maximum temporal level identifier for the i th layer and the maximum temporal level identifier for the (i-1) th layer are determined. Since a large difference rarely occurs, the difference between the maximum time level identifiers of both may not be set to a value between 0 and 7. For example, the difference between the maximum temporal level identifier for the i th layer and the maximum temporal level identifier for the (i-1) th layer may be set within a range of 0 to 3 to be encoded. In this case, the difference time level indicator may have a value within a range of 0 to 3.

Alternatively, when the maximum temporal indicator for the (i-1) th layer has the maximum value within a predetermined range, the value of the differential temporal indicator for the i th layer may be set to zero. This is only allowed when the value of the temporal level identifier is greater than or equal to the lower layer in the upper layer, so that the maximum temporal level identifier for the i th layer is smaller than the maximum temporal level identifier for the (i-1) th layer. Because it is difficult to do.

If the frame rate difference between the i th layer and the (i-1) th layer is large, the difference between the maximum temporal level identifier for the i th layer and the maximum temporal level identifier for the (i-1) th layer is different. Since a large case rarely occurs, there is a high probability that the values of the maximum temporal indicators (max_tid_il_ref_pics_plus1) of all layers are the same. Therefore, the maximum temporal indicator for each layer can be efficiently encoded by using a flag indicating whether the maximum temporal indicator (max_tid_il_ref_pics_plus1) of all layers is the same.

Referring to FIG. 8, a default time level flag (isSame_max_tid_il_ref_pics_flag) for a video sequence may be obtained (S800).

Here, the default temporal flag may mean information indicating whether the maximum temporal indicators (or the maximum temporal identifiers) of all layers in the video sequence are the same.

When the default temporal flag obtained in step S800 indicates that the maximum temporal indicators of all layers in the video sequence are the same, a default maximum temporal indicator (default_max_tid_il_ref_pics_plus1) may be obtained (S810).

Here, the default maximum temporal indicator indicates a maximum temporal indicator that is applied to all layers in common. The maximum temporal level identifier of each layer may be derived from the default maximum temporal indicator, and may be derived, for example, by subtracting 1 from the value of the default maximum temporal indicator.

Alternatively, the default maximum time level indicator may be derived to a pre-defined value. This may be applied when the maximum temporal indicators are not signaled for each layer, such as when the maximum temporal indicators of all the layers in the video sequence are the same. For example, the pre-defined value may mean a maximum value within a predetermined range to which the maximum time level indicator belongs. When the predetermined range for the value of the maximum temporal indicator is 0 to 7, the value of the default maximum temporal indicator may be derived as 7.

On the other hand, when the default temporal flag obtained in step S800 indicates that the maximum temporal indicators of all the layers in the video sequence are not the same, the maximum temporal indicator may be obtained for each layer in the video sequence (S820).

In more detail, the maximum temporal indicator may be obtained by the maximum number of layers allowed in one video sequence. The maximum temporal indicator may be obtained from a video parameter set of the bitstream.

When the value of the obtained maximum temporal indicator is 0, this may mean that the corresponding picture of the lower layer is not used as the interlayer reference picture of the upper layer. In this case, the corresponding picture of the lower layer may be a non-random access picture.

For example, if the value of the maximum temporal indicator is greater than 0, a picture belonging to the i th layer among the plurality of layers of the video sequence and having a temporal level identifier having a value greater than the maximum temporal identifier is (i + 1). It is not used as a reference picture for inter-layer prediction of a picture belonging to the first layer. That is, the (i + 1) th layer only when the value of the maximum temporal indicator is larger than 0 and a picture belonging to the i th layer among the plurality of layers of the video sequence has a temporal level identifier of a value smaller than the maximum temporal identifier. It may be used as a reference picture for inter-layer prediction of a picture belonging to. Here, the maximum temporal level identifier is a value derived from the maximum temporal level indicator. For example, the maximum temporal level identifier may be derived by subtracting 1 from the value of the maximum temporal level indicator.

On the other hand, the maximum time level indicator obtained in step S820 has a value (for example, 0 to 7) within a predetermined range. If the value of the maximum temporal level indicator acquired in step S820 corresponds to the maximum value among the values within the predetermined range, the corresponding picture of the lower layer is higher regardless of the temporal ID (TemporalID) of the corresponding picture of the lower layer. It may be used as an interlayer reference picture of a layer.

If the corresponding picture of the lower layer used as the reference picture for inter-layer prediction of the current picture of the upper layer in the multilayer structure or the picture of the lower layer referring to the corresponding picture of the lower layer is known in advance, other layers in the lower layer may be known. Pictures can be removed from the decoded picture buffer, so that the decoded picture buffer can be managed efficiently. If a picture is not used as an interlayer reference picture or a temporal reference picture, separate signaling may be performed so that the picture is not included in the decoded picture buffer. This is called a discardable flag. Hereinafter, a method of efficiently managing a decoded picture buffer based on the discardable flag will be described with reference to FIG. 9.

9, a discardable flag (discardable_flag) for a picture of a lower layer may be obtained (S900).

The discardable flag may mean information indicating whether a decoded picture is used as a temporal reference picture or an interlayer reference picture in the process of decoding a picture having a lower priority in decoding order. The discardable flag may be obtained in picture units or in slices or slice segments. A detailed method of obtaining the discardable flag will be described with reference to FIGS. 10 to 11.

In operation S910, it may be determined whether a picture of a lower layer is used as a reference picture according to the discardable flag acquired in operation S900.

Specifically, when the discardable flag is 1, it may mean that the encoded picture is not used as the reference picture in the decoding process of the lower priority picture in decoding order. On the other hand, when the discardable flag is 0, it may mean that the hatched picture may be used as a reference picture in the decoding process of a lower priority picture in decoding order.

Here, the reference picture refers to a picture used for inter-layer prediction of a picture of a lower layer and another picture belonging to the same layer (ie, a temporal reference picture) and a picture of a higher layer (ie, an interlayer reference picture). It can be understood as a concept to include.

When the discardable flag indicates that a picture of a lower layer is used as a reference picture in the decoding process of a lower priority picture in decoding order, the picture of the lower layer may be stored in the decoded picture buffer (S920).

In detail, when a picture of a lower layer is used as a temporal reference picture, it may be stored in a decoded picture buffer of the lower layer. When the picture of the lower layer is used as the interlayer reference picture, the picture of the lower layer may further include an upsampling process in consideration of the resolution with the upper layer, and a detailed upsampling process has been described in detail with reference to FIG. 5. The detailed description will be omitted here. The picture of the upsampled lower layer may be stored in the decoded picture buffer of the upper layer.

Meanwhile, when the discardable flag indicates that a picture of a lower layer is not used as a reference picture in the decoding process of a lower priority picture in decoding order, the picture of the lower layer may not be stored in the decoded picture buffer. Alternatively, an identification mark (unused for reference) indicating that the picture or slice is not used as a reference picture for the picture of the lower layer may be marked.

As shown in FIG. 10, a discardable flag (discardable_flag) may be obtained from a slice segment header (S1000).

The slice segment header has only the independent slice segment, and the dependent slice segment may share the slice segment header with the independent slice segment. Therefore, the discardable flag may be obtained in a limited case when the current slice segment corresponds to an independent slice segment.

In FIG. 10, the discardable flag is obtained from the slice segment header. However, the present invention is not limited thereto, and the discardable flag may be acquired in the picture unit or the slice unit.

When the value of the discardable flag acquired in step S1000 is 0, a slice or a picture of a lower layer is used as an interlayer reference picture or another slice of the lower layer in an access unit (AU) including pictures of a multilayer. Or it may be used as a reference picture of the picture. Meanwhile, in order to identify the use as the reference picture, the slice or the picture of the lower layer may be marked as a “short-term reference”.

On the other hand, if the value of the discardable flag acquired in step S1000 is 1, a slice or picture of a lower layer is used as an interlayer reference picture or another layer of the lower layer in an access unit (AU) including pictures of the multilayer. It cannot be used as a reference picture of a slice or picture. Therefore, slices or pictures of the lower layer may be marked with an unused for reference indicating that they are not used as reference pictures.

Alternatively, when the discardable flag has a value of 1, whether a picture of a lower layer is used as an interlayer reference picture or a temporal reference picture in the corresponding access unit may be further considered in consideration of the slice reserved flag (slice_reserved_flag) shown in FIG. 10. You can also decide. Specifically, when the value of the slice preliminary flag is 1, the slice or picture of the lower layer may be set to be used as the interlayer reference picture in the corresponding access unit.

In determining whether the corresponding picture of the lower layer is used as an interlayer reference picture of the current picture of the upper layer, the temporal level identifier TemporalID of the corresponding picture of the lower layer may be considered. That is, as described above with reference to FIG. 3, the corresponding picture of the lower layer may be used as an interlayer reference picture only when the temporal level identifier of the corresponding picture of the lower layer is smaller than or equal to the maximum temporal level identifier of the lower layer. .

As described above, when the temporal level identifier of the corresponding picture of the lower layer is larger than the maximum temporal identifier of the lower layer, the corresponding picture is not used as an interlayer reference picture, so that a discardable flag is not encoded for the corresponding picture. You can not. An identification indication (unused for reference) indicating that the corresponding picture is not used as a reference picture may be marked.

Referring to FIG. 11, it may be compared whether a temporal ID (TemporalID) of a picture or slice belonging to a lower layer is equal to, smaller, or larger than the maximum temporal identifier (max_tid_il_ref_pics [nuh_layer_id-1]) of the lower layer (S1100). ).

As a result of the comparison in step S1100, the discardable flag (discardable_flag) only when the temporal ID (TemporalID) of the picture or slice belonging to the lower layer is equal to or smaller than the maximum temporal identifier (max_tid_il_ref_pics [nuh_layer_id-1]) of the lower layer. ) Can be obtained (S1110).

On the other hand, when the value of the discardable flag acquired in step S1110 is 1 or the temporal ID (TemporalID) of the picture or slice belonging to the lower layer is larger than the maximum temporal level identifier (max_tid_il_ref_pics [nuh_layer_id-1]) of the lower layer. Since the picture or the slice of the lower layer is not used as the reference picture, an identification mark for this can be marked.

On the other hand, the temporal ID (TemporalID) of the picture or slice belonging to the lower layer is equal to or smaller than the maximum temporal identifier (max_tid_il_ref_pics [nuh_layer_id-1]) of the lower layer, and the value of the discardable flag acquired in step S1110 is 0. In the case of, since the picture or slice of the lower layer may be used as the reference picture, this may mark a short-term reference.

The present invention can be used to code a scalable video signal.

Claims

Obtaining a discardable flag for a picture of a lower layer;

Determining whether a picture of the lower layer is used as a reference picture based on the discardable flag; And

When the picture of the lower layer is used as a reference picture, the picture of the lower layer is stored in the decoded picture buffer,

And the discardable flag is information indicating whether a decoded picture is used as the reference picture in decoding a picture having a lower priority in decoding order.
The method of claim 1, wherein the discardable flag is obtained from a slice segment header.
The method of claim 2, wherein the discardable flag is obtained when a temporal level identifier of a picture of the lower layer is less than or equal to a maximum temporal level identifier for the lower layer.
The method of claim 1, wherein the stored lower layer picture is marked as a short-term reference picture.
An entropy decoding unit for obtaining a discardable flag for a picture of a lower layer; And

A decoded picture buffer for determining whether a picture of the lower layer is used as a reference picture based on the discardable flag, and if a picture of the lower layer is used as a reference picture, decoded picture buffer for storing the picture of the lower layer. Including,

And the discardable flag is information indicating whether a decoded picture is used as the reference picture in decoding a picture having a lower priority in decoding order.
6. The scalable video signal decoding apparatus of claim 5, wherein the discardable flag is obtained from a slice segment header.
The apparatus of claim 6, wherein the discardable flag is obtained when a temporal level identifier of a picture of the lower layer is equal to or smaller than a maximum temporal level identifier for the lower layer.
The apparatus of claim 5, wherein the stored lower layer picture is marked as a short-term reference picture.
Obtaining a discardable flag for a picture of a lower layer;

Determining whether a picture of the lower layer is used as a reference picture based on the discardable flag; And

When the picture of the lower layer is used as a reference picture, the picture of the lower layer is stored in the decoded picture buffer,

And the discardable flag is information indicating whether a decoded picture is used as the reference picture in decoding a picture having a lower priority in decoding order.
10. The method of claim 9, wherein the discardable flag is obtained from a slice segment header.
12. The method of claim 10, wherein the discardable flag is obtained when a temporal level identifier of a picture of the lower layer is equal to or smaller than a maximum temporal level identifier for the lower layer.
10. The method of claim 9, wherein the stored lower layer picture is marked as a short-term reference picture.
An entropy decoding unit for obtaining a discardable flag for a picture of a lower layer; And

A decoded picture buffer for determining whether a picture of the lower layer is used as a reference picture based on the discardable flag, and if a picture of the lower layer is used as a reference picture, decoded picture buffer for storing the picture of the lower layer. Including,

And the discardable flag is information indicating whether a decoded picture is used as the reference picture in decoding a picture having a lower priority in decoding order.
14. The scalable video signal encoding apparatus of claim 13, wherein the discardable flag is obtained from a slice segment header.
15. The apparatus of claim 14, wherein the discardable flag is obtained when a temporal level identifier of a picture of the lower layer is equal to or smaller than a maximum temporal level identifier for the lower layer.